# US 12,602,626 B2 · Systems and methods for generating time slot predictions and repurchase predictions using machine learning architectures > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 12,602,626 B2 - Document ID: US12602626B2 - Status: Granted - Kind: B2 - Application number: 18/425,114 - Filed: 2024-01-29 - Published: 2024-05-23 - Granted: 2026-04-14 - Earliest claimed: 2021-01-30 - Assignees: Walmart Apollo, LLC - Applicants: Walmart Apollo, LLC - Inventors: Sonal Bathe; Rahul Sridhar; Sinduja Subramaniam; Evren Korpeoglu; Kannan Achan - Web page: https://patentgrep.com/patent/US12602626B2 ## CPC - G06Q 10/02: Reservations, e.g. for tickets, services or events - G06N 20/00: Machine learning - G06Q 10/04: Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" (market predictions or forecasting for commercial activities G06Q30/0202) - G06Q 10/0836: Recipient pick-ups - G06Q 10/1093: Calendar-based scheduling for persons or groups - G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound - G06N 20/20: Ensemble learning - G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS ## Abstract A system including one or more processors and one or more non-transitory computer-readable storage devices storing computing instructions that, when executed on the one or more processors, cause the one or more processors to perform operations: generating, using a first machine learning model, a first output comprising a repurchase prediction for a user; generating, using a second machine learning model and using respective data of the repurchase prediction of the first machine learning model for the user, a second output comprising a time slot prediction for the user; initiating one or more reservation functions based at least in part on the first output and the second output; and transmitting an option to the user to access a GUI of a digital shopping cart system to reserve a reservation function of the one or more reservation functions. Other embodiments are described. ## Claims ### Claim 1 A system comprising: one or more processors; and one or more non-transitory computer-readable storage devices storing computing instructions that, when executed on the one or more processors, cause the one or more processors to: train a gradient boosting model by providing, to the gradient boosting model, one or more feature vectors, that are based on data associated with previous transactions conducted on an electronic platform, and corresponding labels to teach the gradient boosting model a sequence of decision trees via a boosting method; generate, using a first machine learning model of a machine learning architecture, that includes the gradient boosting model, a first output comprising a repurchase prediction for a user by computing, during inference, a prediction of each tree, of the sequence of decision trees, based on a feature vector of the one or more feature vectors to reduce a prediction error before outputting a class, for the repurchase prediction, indicating whether a transaction will occur, on the electronic platform, in a subsequent period of time; generate, using a second machine learning model of the machine learning architecture and using respective data of the repurchase prediction of the first machine learning model for the user, a second output comprising a time slot prediction for the user, wherein the time slot prediction predicts a time slot for the user for an upcoming transaction, and wherein the first machine learning model is performed independently from the second machine learning model; initiate one or more reservation functions based at least in part on the first output and the second output; integrate the one or more reservation functions with a digital shopping cart system to create the upcoming transaction based on the repurchase prediction and the time slot prediction previously generated by the machine learning architecture; and transmit an option to access a GUI of the digital shopping cart system to reserve a reservation function of the one or more reservation functions. ### Claim 2 (depends on 1) The system of claim 1, wherein: the repurchase prediction estimates a probability that the user at least (i) repurchases one or more items within a time period, or (ii) conducts a transaction on an electronic platform within the time period; and the time slot prediction is generated in response to the repurchase prediction indicating the user is likely to at least (i) repurchase the one or more items, or (ii) conduct the transaction on the electronic platform within the time period. ### Claim 3 (depends on 1) The system of claim 1, wherein the computing instructions, when executed on the one or more processors, further cause the one or more processors to: execute the one or more reservation functions to facilitate reserving the time slot for the upcoming transaction for the user. ### Claim 4 (depends on 1) The system of claim 1, wherein at least one of the first machine learning model or the second machine learning model comprises a random forest decision tree model that is configured to generate at least one of: the repurchase prediction; or the time slot prediction. ### Claim 5 (depends on 1) The system of claim 1, wherein at least one of the first machine learning model or the second machine learning model comprises a logistic regression model configured to generate at least one of: the repurchase prediction; or the time slot prediction. ### Claim 6 (depends on 1) The system of claim 1, wherein the reservation function is configured to transmit a notification to the user which notifies the user that a time slot is available. ### Claim 7 (depends on 1) The system of claim 1, wherein the reservation function is integrated or communicates with an electronic platform that offers items over a network. ### Claim 8 (depends on 1) The system of claim 1, wherein the computing instructions, when executed on the one or more processors, further cause the one or more processors to: configure at least one of the first machine learning model or the second machine learning model based on at least one of: a frequency of the user conducting transactions; a date or a time the user conducted a transaction on the electronic platform; one or more of a mean time or a median time between transactions conducted by the user; or a number of days that elapsed since a most-recent transaction of the user. ### Claim 9 (depends on 1) The system of claim 1, wherein the one or more feature vectors include transaction-based features extracted or derived from the data about the user. ### Claim 10 (depends on 9) The system of claim 9, wherein the one or more feature vectors include slot-based features associated with one or more time slots that were previously selected by the user in connection with the previous transactions. ### Claim 11 A method, comprising: generating, using a first machine learning model of a machine learning architecture, that includes a gradient boosting model, a first output comprising a repurchase prediction for a user by computing, during inference, a prediction of each tree, of a sequence of decision trees, based on a feature vector before outputting a class, for the repurchase prediction, indicating what will occur, on an electronic platform, in a subsequent period of time, wherein the gradient boosting model is trained through providing, to the gradient boosting model, one or more feature vectors, that are based on data associated with previous transactions conducted on the electronic platform, and corresponding labels to teach the gradient boosting model a sequence of decision trees via a boosting method, and wherein the one or more feature vectors include the feature vector; generating, using a second machine learning model of the machine learning architecture and using respective data of the repurchase prediction of the first machine learning model for the user, a second output comprising a time slot prediction for the user, wherein the time slot prediction predicts a time slot for the user for an upcoming transaction, and wherein the first machine learning model is performed independently from the second machine learning model; and initiating one or more reservation functions based at least in part on one or more of the first output or the second output. ### Claim 12 (depends on 11) The method of claim 11, wherein: the repurchase prediction estimates a probability that the user at least (i) repurchases one or more items within a time period, or (ii) conducts a transaction on an electronic platform within the time period; and the time slot prediction is generated in response to the repurchase prediction indicating the user is likely to at least (i) repurchase the one or more items, or (ii) conduct the transaction on the electronic platform within the time period. ### Claim 13 (depends on 11) The method of claim 11, further comprising: executing the one or more reservation functions to facilitate reserving of the time slot for the upcoming transaction for the user. ### Claim 14 (depends on 11) The method of claim 11, wherein at least one of the first machine learning model or the second machine learning model comprises a random forest decision tree model that is configured to generate at least one of: the repurchase prediction; or the time slot prediction. ### Claim 15 (depends on 11) The method of claim 11, wherein at least one of the first machine learning model or the second machine learning model comprises a logistic regression model or gradient boosted tree model configured to generate at least one of: the repurchase prediction; or the time slot prediction. ### Claim 16 (depends on 11) The method of claim 11, wherein the one or more reservation functions include a reservation function that is configured to transmit a notification that the time slot is available. ### Claim 17 (depends on 11) The method of claim 11, wherein the one or more reservation functions include a reservation function that is integrated or communicates with the electronic platform. ### Claim 18 (depends on 11) The method of claim 11, wherein the data is based on extracted feature vectors, and wherein the extracted feature vectors comprise transaction-based features and slot-based features over a period of time. ### Claim 19 A non-transitory computer-readable medium storing instructions, the instructions, upon execution by a processor, cause the processor to perform operations comprising: training a gradient boosting model by providing, to the gradient boosting model, one or more feature vectors, that are based on data associated with previous transactions conducted on an electronic platform, and corresponding labels to teach the gradient boosting model a sequence of decision trees via a boosting method; generating, using a first machine learning model of a machine learning architecture, that includes the gradient boosting model, a first output comprising a repurchase prediction by computing a prediction of each tree, of the sequence of decision trees, based on a feature vector of the one or more feature vectors before outputting a class, for the repurchase prediction, indicating what will occur, on the electronic platform, in a subsequent period of time; generating, using a second machine learning model of the machine learning architecture and using respective data of the repurchase prediction of the first machine learning model, a second output comprising a time slot prediction for a user, wherein the time slot prediction predicts a time slot for the user for an upcoming transaction; and initiating one or more reservation functions based at least in part on the second output. ### Claim 20 (depends on 19) The non-transitory computer-readable medium storing the instructions of claim 19, wherein the instructions, upon execution by the processor, further cause the processor to perform operations comprising: integrating the one or more reservation functions with a digital shopping cart system to create the upcoming transaction based on the repurchase prediction and the time slot prediction previously generated by the machine learning architecture; and executing the one or more reservation functions to facilitate reserving the time slot for the upcoming transaction for the user.